jameshfisher

package trie

import (
	"container/vector"
	"sort"
)

// A 'set' structure, that can hold []byte objects.
// For any one []byte instance, it is either in the set or not.
type Trie struct {

spicycode

# 0 is too far from ` ;)
set -g base-index 1

# Automatically set window title
set-window-option -g automatic-rename on
set-option -g set-titles on

#set -g default-terminal screen-256color
set -g status-keys vi
set -g history-limit 10000

erikreagan

Mac web developer apps

This gist's comment stream is a collection of webdev apps for OS X. Feel free to add links to apps you like, just make sure you add some context to what it does — either from the creator's website or your own thoughts.

— Erik

naholyr

Sample service script for debianoids

Look at LSB init scripts for more information.

Usage

Copy to /etc/init.d:

# replace "$YOUR_SERVICE_NAME" with your service's name (whenever it's not enough obvious)

rokirx

Functional Programming Resources

Functional Programming in general

http://www.youtube.com/watch?v=1PhArSujR_A&t=2m5s

http://www.altdevblogaday.com/2012/04/26/functional-programming-in-c/

Purely functional priority queues

bjorgvino

Remove osxfuse if installed via homebrew:
> brew uninstall osxfuse

Install osxfuse binary and choose to install the MacFUSE compatibility layer:
http://sourceforge.net/projects/osxfuse/files/latest/download?source=files

Reboot (optional but recommended by osxfuse)

Install ntfs-3g via homebrew:
> brew update && brew install ntfs-3g

JeffBelback

#!/bin/bash
# Stop all containers
containers=`docker ps -a -q`
if [ -n "$containers" ] ; then
        docker stop $containers
fi
# Delete all containers
containers=`docker ps -a -q`
if [ -n "$containers" ]; then
        docker rm -f -v $containers

yoavg

Reinforcement Learning for Language Models

Yoav Goldberg, April 2023.

Why RL?

With the release of the ChatGPT model and followup large language models (LLMs), there was a lot of discussion of the importance of "RLHF training", that is, "reinforcement learning from human feedback". I was puzzled for a while as to why RL (Reinforcement Learning) is better than learning from demonstrations (a.k.a supervised learning) for training language models. Shouldn't learning from demonstrations (or, in language model terminology "instruction fine tuning", learning to immitate human written answers) be sufficient? I came up with a theoretical argument that was somewhat convincing. But I came to realize there is an additional argumment which not only supports the case of RL training, but also requires it, in particular for models like ChatGPT. This additional argument is spelled out in (the first half of) a talk by John Schulman from OpenAI. This post pretty much